Left ventricular reshaping: Effects on the pressure-volume relationship

Left ventricular reshaping: Effects on the pressure-volumerelationshipAbul Kashem, MD, PhDa

Sarmina Hassan, MD, PhDa

Deborah L. Crabbe, MDa

David B. Melvin, MDb

William P. Santamore, PhDa

Dr Kashem

Objective: We tested whether the CardioClasp device (CardioClasp, Inc, Cincinnati,Ohio), a non-blood contact device, would improve left ventricular contractility byacutely reshaping the left ventricle and reducing left ventricular wall stress.

Methods: In dogs (n � 6) 4 weeks of ventricular pacing (210-240 ppm) inducedsevere heart failure. Left ventricular function was evaluated before and afterplacement of the CardioClasp device, which uses 2 indenting bars to reshape the leftventricle. Hemodynamics, echocardiography, and Sonometrics crystals dimension(Sonometrics Corporation, London, Ontario, Canada) were measured at steady stateand during inferior vena caval occlusion.

Results: The CardioClasp device decreased the left ventricular end-diastolic ante-rior-posterior dimension by 22.8% � 1.9%, decreased left ventricular wall stressfrom 97.3 � 22.8 to 67.2 � 7.7 g/cm2 (P � .003), and increased the fractional areaof contraction from 21.3% � 10.5% to 31.3% � 18.1% (P � .002). The clasp didnot alter left ventricular end-diastolic pressure, left ventricular pressure, left ven-tricular dP/dt, or cardiac output. With the CardioClasp device, the slope of theend-systolic pressure-volume relationship was increased from 1.87 � 0.47 to 3.22 �1.55 mm Hg/mL (P � .02), the slope of preload recruitable stroke work versusend-diastolic volume was increased from 28.4 � 11.0 to 44.1 � 23.5 mm Hg (P �.02), and the slope of maximum dP/dt versus end-diastolic volume was increasedfrom 10.6 � 4.6 to 18.6 � 7.4 mm Hg�s�1�mL�1 (P � .01). The CardioClaspdevice increased the slope of the end-systolic pressure-volume relationship by68.0% � 21.7%, the slope of preload recruitable stroke work versus end-diastolicvolume by 50.7% � 18.1%, and the slope of maximum dP/dt versus end-diastolicvolume by 85.7% � 28.9%.

Conclusions: The CardioClasp device decreased left ventricular wall stress andincreased the fractional area of contraction by reshaping the left ventricle. TheCardioClasp device was able to maintain cardiac output and arterial pressure. Theclasp increased global left ventricular contractility by increasing the slope of theend-systolic pressure-volume relationship, the slope of preload recruitable strokework versus end-diastolic volume, and the slope of maximum dP/dt versus end-diastolic volume. In patients with heart failure, the CardioClasp device might beeffective for clinical application.

With heart failure, left ventricular (LV) enlargement places anincreased wall tension on the individual myocytes. Thisincreased wall tension, together with cardiomyocyte dys-function, leads to a profound depression in global LV func-tion and contractility. If this extra geometric burden causedby ventricular dilatation can be eliminated, myocardial wall

stress would decrease, leading to improved LV systolic ventricular performance.

From Cardiovascular Research, TempleUniversity, Philadelphia, Pa,a and the De-partment of Surgery, University of Cincin-nati, Cincinnati, Ohio.b

Supported by a grant from CardioClasp,Inc.

David B. Melvin, MD, is the inventor ofCardioClasp device and owns indirect eq-uity in CardioClasp, Inc, Cincinnati, Ohio;William P. Santamore, PhD, is the Consul-tant for the CardioClasp, Inc; and AbulKashem, MD, PhD, is the institutional prin-cipal investigator for the CardioClasp de-vice.

Read at the Eighty-second Annual Meetingof The American Association of ThoracicSurgery, Washington, DC, May 5-8, 2002.

Received for publication June 4, 2002; re-visions requested July 22, 2002; revisionsreceived July 31, 2002; accepted for publi-cation Aug 2, 2002.

Address for reprints: Abul Kashem, MD,PhD, Temple University School of Medi-cine, Medical Research Building, Room800A, 3420 N. Broad St, Philadelphia, PA19140 (E-mail: [email protected]).

J Thorac Cardiovasc Surg 2003;125:391-9

Copyright © 2003 by The American Asso-ciation for Thoracic Surgery

0022-5223/2003 $30.00�0

doi:10.1067/mtc.2003.4

Kashem et al Evolving Technology

The Journal of Thoracic and Cardiovascular Surgery ● Volume 125, Number 2 391

ET

Melvin1 was the first to examine the concept of passiveventricular remodeling to reduce this geometric burden.Using a computational analysis, he showed that by impos-ing a shape change on a dilated left ventricle, the ratio ofwall tension to chamber pressure was reduced. On the basisof this concept, the CardioClasp device (CardioClasp, Inc,Cincinnati, Ohio) was developed. The CardioClasp deviceis a passive cardiac support device (CSD) that uses 2 in-denting bars to reshape the left ventricle as 2 widely com-municating lobes of reduced radius.

By passively reshaping the left ventricle and decreasingthe radius of curvature, the CardioClasp device reduces LVwall stress.1 In previous studies passive reshaping has beenshown to increases LV ejection fraction without alteringcardiac output or systolic pressures.2-9 In the present study,using a dilated heart failure model, we tested whether theCardioClasp, by passively reshaping the left ventricle andreducing LV wall stress, can improve global LV contractil-ity.

MethodsAll study procedures were performed in compliance with the“Principles of Laboratory Animal Care” formulated by the Na-tional Society for Medical Research and the “Guide for Care andUse of Laboratory Animals” issued by the National Academy ofSciences and published by the National Institutes of Health (Na-tional Institutes of Health publication, revised 1996).

Induction of Heart FailureAfter an overnight fast, 6 mongrel dogs (22.5-26.7 kg) wereanesthetized with intravenous sodium thiopental (15-25 mg/kg)and intramuscular atropine (0.01 mg/kg), intubated, and mechan-ically ventilated (model R-DRAGERAV, S/N. R-2035, NorthAmerican Drager, Telford, Pa). Anesthesia was maintained with1% to 2% isoflurane (Isoflurane Vaporizer, Oharda; Isotec 3,Aushell, Ga) and oxygen. Surface electrocardiographic results(model 78346A; Hewlett Packard, Palo Alto, Calif) and oxygensaturation were monitored continuously.

As described previously, a bipolar, implantable, screw-in,transvenous myocardial pacing lead (model 4058; Medtronic, Min-neapolis, Minn) was introduced through the right jugular vein andpositioned at the right ventricular apex during fluoroscopy.2,10 Thepacing lead was tunneled to the back of the neck and connected toa pulse generator (THERAs model 8966i B, Medtronic) implantedsubcutaneously through a 4- to 5-cm incision. The pacemakergenerator was secured with the underlying muscle, and all thewounds were closed in layers.

Rapid right ventricular pacing was started on the second post-operative day. During the first week, pacing parameters were 210ppm, 3.5-V pulse amplitude, 2.8-mV pulse sensitivity, and 1.5-mspulse width. The pacing rate was increased consecutively to 220,230, and 240 ppm during the second, third, and fourth weeks,respectively.2,10 As in previous studies, incremental pacing reli-ably induced congestive heart failure manifested by clinical find-ings of ascites, muscle wasting, lethargy, and decreased appetite.

Evaluation of the CardioClasp DeviceThe heart failure animals were anesthetized with a combination ofintravenous injections of diazepam (1 mg/kg), fentanyl citrate (2-8�g/kg), and 2% lidocaine (1 mg/kg). The animals were ventilatedthrough an endotracheal tube with a positive-pressure respirator(North American Dragger, model R-DRAGERAV, S/N. R-2035)and maintained with 1% to 2% isoflurane. The electrocardiogramwas monitored continuously. An oxygen saturation monitor(52000 series, S/N. 9801168, WelchAllyn; Tycos Instruments Inc,Ithaca, NY) to measure the oxygen saturation was used continu-ously.

After administering 3000 units of heparin, a 6F micromanom-eter-tipped pigtail catheter (Millar Inc, Houston, Tex) was insertedand advanced to the left ventricle for measurement of the LVpressure. The side port of the catheter sheath was connected to afluid transducer for measuring the arterial pressure. With theanimal in the right lateral position, a left lateral thoracotomy wasperformed at the fifth intercostal space. The pericardium wasopened by using a fine scissor, and the heart was suspended in apericardial cradle. A flow probe (A-series 16-mm, S/L no.16A320;Transonic Systems Inc, Ithaca, NY) was placed around the ascend-ing aorta and connected to a flowmeter (model T206, S/L no.1206-S-991539, Transonic Systems Inc) to measure aortic flow.

A balloon occlusion catheter (8F, 40-mm balloon; Boston Sci-entific Corp, Natick, Mass) introduced through the right femoralvein was placed at the junction of the inferior vena cava (IVC) andright atrium. When required in the protocol, inflating this balloonobstructed flow in the IVC, thereby gradually decreasing cardiacoutput.

Sonometric crystals placement. Six small piezo electric Sono-metrics crystals (Sonomicrometer Crystals; Sonometrics Inc, On-tario, Canada) were placed into the endocardial positions in theanterior, posterior, apex, base, free, and septal walls to assess LVvolume (end-systolic pressure-volume relationship [ESPVR]).Crystal dimensions were measured continuously (SonometricsData Acquisition System, Sonometrics Inc).

Description of the CardioClasp device. The CardioClasp is animplantable medical device that consists of 3 primary components:2 rigid bars with pads and an adjustable tether. The rigid bars havea defined curvature similar to that of the heart. Longitudinally andtorsionally flexible contact pads are attached to each bar. Anadjustable tether connects the 2 pad-bar assemblies on either sideof the left heart.

Placement of the CardioClasp device. All pulmonary veinswere mobilized from the superior to inferior aspects, and a guidingcatheter was introduced underneath the pulmonary veins. Theposterior bar of the CardioClasp was positioned along the long axisof the left ventricle adjacent to the posterior descending artery, andthe anterior bar of the device was placed along the long axis of theleft ventricle adjacent to the left anterior descending coronaryartery. The anterior and posterior bars were connected with a tetherto an adjustment tool. The device, once implanted, was secured tothe left heart by means of mechanical fixation. Once the devicewas fixed, the adjustment tool pulled the tether, shortening thedistance between the anterior and posterior bars and thereby de-creasing the LV anterior posterior (A-P) dimension. We measureda baseline A-P end-diastolic dimension at the papillary musclelevel in a short-axis echocardiographic view. After tightening the

Evolving Technology Kashem et al

392 The Journal of Thoracic and Cardiovascular Surgery ● February 2003

ET

CardioClasp with the adjustment tool, the percentage of A-Pdimension reduction was confirmed by using the short-axis echo-cardiographic view at the papillary muscle level.

MeasurementsAt baseline, the echocardiographic results, arterial and LV pres-sure, and aortic flow were measured, recorded, and digitized on arecorder (VIPER, Gould, Ohio). With the respirator off, thesemeasurements were obtained under steady-state conditions andduring IVC occlusions. In each experiment 3 IVC runs at baselinewere recorded.

Echocardiographic measurements were obtained by using anHP Sonos 5500 machine (Hewlett Packard). All studies wereperformed on animals with an open chest and using an epicardialwindow. An s12 multifrequency transducer (Hewlett Packard) setat the midrange frequency was placed over the LV free wall byusing a stand-off gel pad to obtain images. Fine epicardial suturesfor repositioning marked the probe position. Two-dimensionalshort-axis images of the left ventricle were obtained at the level ofthe papillary muscle. LV internal dimensions were measured fromthese 2-dimensional images. End-diastolic measurements wereobtained at the peak of the R wave on the electrocardiogram.End-systolic measurements were obtained at the end of the Twave.

After obtaining the baseline measurements, the CardioClaspwas placed on the heart, and the LV A-P internal end-diastolicdimension was reduced by approximately 25%. Echocardiogramswere obtained. Hemodynamic data were acquired with the Cardio-Clasp. The echocardiographic results, arterial and LV pressure,and aortic flow were recorded with the respirator off. These mea-surements were obtained under steady-state conditions and duringIVC occlusions similar to the baseline measurements.

Finally, the animal was killed with 2 mEq/kg potassium chlo-ride (20-30 mL of KCl) administered intravenously and 15 mg/kgsodium pentothal (10-15 mL) administered intravenously.

Data AnalysisEchocardiography. The echocardiographic images at end sys-

tole and end diastole were captured by the Hewlett-Packard Echo-cardiography machine (Sonos 5500), where the LV end-diastolicand end-systolic areas’ (short-axis view) endocardial borders weretraced. The dimensional and area information were used to calcu-late the area ejection fraction or fractional area of contraction(FAC) as follows:

FAC � (LVEDA–LVESA)/LVEDA � 100

where LVDEA is defined as LV end-diastolic area, and LVESA isdefined as LV end-systolic area.

Peak LV wall stress was calculated as follows:

LVWS � (0.334 � LVP � LVD)/[WT (1 � WT/LVD)]

where LVP is defined as LV systolic pressure, LVD is defined asLV anterior posterior diameter, and WT is defined as LV wallthickness.11,12

Hemodynamics. By using software developed in Visual Basicfor Excel (Microsoft Excel 7.0; Microsoft Inc, Redwood, Wash),hemodynamic variables were extracted from a digitally stored data

file. Ectopic and postectopic cycles were excluded from the anal-ysis. For each cardiac cycle, the end-diastolic pressure, the peakventricular systolic pressure, LV �dP/dt, LV �dP/dt, and peakand end-diastolic aortic pressures were determined, and strokevolume and work were calculated.

Indices of contractility. Multiple LV pressure-volume loopswere obtained during transient preload reduction induced by oc-cluding the IVC. Premature beats and subsequent beat were ex-cluded from the analysis. LV contractility was assessed by theend-systolic pressure-volume (PES-VES) relationship (PES �EES[VES�VO], where EES is the slope of the ESPVR)13, thepreload recruitable stroke work (SW) versus end-diastolic volume(SW � MSW [VED�VO,SW]) relationship (MSW),14 and the max-imum dP/dt versus VED (dP/dtMAX � dE/dtMAX[VED�VO,SW])relationship.15

Data are expressed as means � SD. The hemodynamic, echo-cardiographic, and indices data at baseline were compared withthose during use of the CardioClasp by using the paired Student ttest.

ResultsFigure 1 shows 2 representative LV short-axis echocardio-graphic views at the papillary muscle level at baseline andwith the CardioClasp. The LV short-axis views were ob-tained at end diastole. With the CardioClasp, the LV A-Pdiameter was reduced by 25%, from 4.56 to 3.43 cm.Fractional areas of contractions were increased by 22.0%. Inthe group data echocardiographic images revealed that theCardioClasp device reduced the end-diastolic LV A-P di-mension by 22.8% � 1.9%. Associated with this shapechange, the CardioClasp decreased LV end-diastolic area by20.4% � 5.0% and LV end-systolic area by 25.1% � 9.5%.

Table 1 presents the absolute values for echocardio-graphic parameters at baseline and with the CardioClasp.With the CardioClasp, end-systolic and end-diastolic LVA-P and septal free wall thickness were increased slightly.End-diastolic septal free wall dimension were unchanged,whereas end-systolic septal free wall dimensions were in-creased (P � .022).

With the CardioClasp (Figure 2), myocardial LV wallstress decreased significantly from baseline values (97.3 �22.8 to 67.2 � 7.7 g/cm2, P � .003). The CardioClaspdecreased myocardial LV wall stress by 28.9% � 4.4%compared with baseline values.

With the CardioClasp (Figure 3), the FAC increasedsignificantly from baseline (21.3 � 10.5 to 31.3 � 18.1,P � .002). The CardioClasp increased FAC significantly by50.9% � 18.1% compared with baseline values.

From one experiment, Figure 4 shows different indicesof LV contractility. In Figure 4, A, LV pressure-volumeloops are plotted at baseline and with the CardioClasptogether with the ESPVRs. In this example the CardioClaspincreased the slope of the ESPVR. Figure 4, A, shows oneexample of ESPVR loops generated during the baselineperiod and with the CardioClasp at 25%. The data show a



ET

steepening of the LV ESPVR slope (increased Emax) at25%.

Figure 4, B, shows the effects of the CardioClasp onpreload recruitable stroke work. The rounded bullets andsolid line represent the SW-EDV relationship with the Car-dioClasp and the dotted line with solid bullets represents theSW-EDV relationship at baseline. Clearly, the CardioClaspincreased the SW-EDV relationship compared with baselinevalues.

Figure 4, C, shows the effects of the CardioClasp ondE/dt, the slope of the maximum dP/dt–VED(dP/dtMAX�VED) relationship. The rounded bullets and solidline represent dE/dt with the CardioClasp, and the dotted

line with solid bullets represents dE/dt at baseline. TheCardioClasp increased dE/dt compared with baseline val-ues.

Table 2 presents the absolute hemodynamic parametersat baseline and with the CardioClasp. With the CardioClasp,cardiac output was unaltered (P � .40). Likewise, peak LVsystolic (P � .26) and LV end-diastolic (P � .43) pressures,peak positive (P � .12) and negative (P � .23) first deriv-atives of LV pressures (dP/dts), and aortic systolic (P � .38)and diastolic (P � .12) pressures were unaltered.

LV contractility, assessed by means of EES, MSW, anddE/dtMAX, was significantly greater in the CardioClasp-treated group than in the baseline heart failure group. Table

Figure 1. Two representative LV end-diastolic short-axis echocardiographic views at papillary muscle level atbaseline heart failure and with the CardioClasp. The CardioClasp significantly reduced LV A-P diameter by 25%.The figure shows how the A-P dimension was measured connecting the line immediately below to the anterior andposterior papillary muscles of the left ventricle (A, corner point of the anterior papillary muscle; P, corner pointof the posterior papillary muscle). In this image the anterior bar of the CardioClasp device was in front of theanterior papillary muscle and not visible. The posterior bar of the CardioClasp device was at the back of theposterior papillary muscle and poorly visible (slightly echogenic behind the papillary muscle).

TABLE 1. Effects of the CardioClasp on echocardiographic parametersLV echocardiographic parameters Baseline CardioClasp P value

Diastole A-P dimension (cm) 4.31 � 0.34 3.39 � 0.34* .0001S-FW dimension (cm) 4.04 � 0.34 4.04 � 0.42 .488A-P wall thickness (mm) 9.6 � 2.1 9.7 � 1.3 .431S-FW wall thickness (mm) 9.4 � 1.7 9.5 � 1.1 .485End-diastolic area (cm2) 12.2 � 2.1 9.7 � 2.3* .007

Systole A-P diameter (cm) 3.73 � 0.36 3.10 � 0.44* .0001S-FW dimension (cm) 3.70 � 0.55 3.92 � 0.54* .022A-P wall thickness (mm) 10.0 � 1.7 11.1 � 2.2* .040S-FW wall thickness (mm) 10.1 � 1.6 11.3 � 2.2 .052End-systolic area (cm2) 10.8 � 2.6 8.0 � 3.1* .034

S-FW, Septal free wall.*P � .05.



ET

3 shows the systolic indices of the pressure-volume rela-tionship at baseline and with the CardioClasp. With theCardioClasp, Emax increased by 68.0% � 21.7% (P � .02),preload recruitable stroke work increased by 50.7% �18.1% (P � .02), and dE/dt increased by 85.7% � 28.9%(P � .01).

DiscussionThe concept and understanding of passive LV geometricreshaping is evolving. Melvin1 was the first to examine theconcept of passive geometric ventricular remodeling byusing a computational analysis. The model showed that byimposing a shape change on a dilated left ventricle, the ratioof wall tension to chamber pressure was reduced. If thereduction in wall tension occurred as a result of ventricularresection, then a substantial degree of contractile shorteningimprovement would be needed just to maintain baselinestroke volume.

Shimizu and colleagues3 examined the effects of passiveLV reshaping in a canine model of dilated heart failure.After rapid ventricular pacing induced heart failure, theheart was attached to an isolated heart preparation. TheCardioClasp was positioned on the heart and adjusted toreduce A-P end-diastolic dimension by 20% to 30%. Withthe CardioClasp, the ESPVR was significantly increasedfrom 1.4 mm Hg/mL at baseline to 2.4 mm Hg/mL. Aftertaking off the device, ESPVR returned to the baseline level(1.5 mm Hg/mL). Shimizu and colleagues also showed thatboth the systolic and diastolic pressures and the developedpressure decreased with rapid removal of the CardioClasp.

Earlier, our group reported acute hemodynamic improve-ments with the CardioClasp at approximately 30% and 15%LV end-diastolic dimension reduction.2 We showed LVdimension reduction at 31.3%, and the FAC was increasedby 27%.

Santamore and coworkers4 reported that at normal after-load, the CardioClasp maintained the stroke volume versus

end-diastolic pressure relationship. At increased afterload,the decrease in the stroke volume versus end-diastolic pres-sure relationship was less with the CardioClasp. Santamoreand coworkers also reported that LV reshaping with theCardioClasp attenuated the cardiac output decrease in re-sponse to increased afterload. With increased afterload, thestroke volume decreased significantly less with the Cardio-Clasp (�20.4% � 0.4%) compared with control values(�27.7% � 0.6%, P � .05).

In our experiments the CardioClasp acutely reduced theLV A-P diameter by 22.8%, as demonstrated by means ofdirect epicardial transthoracic echocardiography. The Car-dioClasp device also significantly reduced the LV systolicwall stress by 29%. Our measurement of FAC clearlyshowed the effective results with the passive CardioClaspdevice; the FAC was increased by 51% with the Cardio-Clasp at implantation. In these experiments the CardioClaspincreased Emax by 68.0%, Msw by 50.7%, and dE/dtMAX by85.7%, indicating improved LV myocardial contractility.

Baseline LV systolic and diastolic pressures were notsignificantly altered compared with pressures with use ofthe CardioClasp. Likewise, cardiac output was unalteredwith use of the CardioClasp at implantation (1.9 L/min)compared with baseline heart failure. The reason for in-creased ejection fraction and contractility with unalteredcardiac output (1.9 L/min) probably lies in the decreased LVend-diastolic volume, as evidenced by the decreased end-diastolic areas after CardioClasp placement.

The key findings of the present study are the following:1. The CardioClasp reshaped the left ventricle, thereby

reducing the LV systolic wall stress.2. The FAC increased, whereas cardiac output and LV

pressure were unaltered.3. The CardioClasp significantly increased global LV

contractility, as indicated by the increases in Emax,MSW, and the slope of the maximum dP/dt–VED(dE/dtMAX) relationship.

Figure 3. Significantly increased LV FAC by 50.9% � 18.1% withthe CardioClasp. *P < .05.

Figure 2. Significantly decreased LV wall stress (LVWS) with theCardioClasp by 28.9% � 4.4%. *P < .05.



ET

Comparison with the LiteratureThe Myosplint device pulls the opposing walls of the hearttogether, thereby decreasing the chamber radius. Each Myo-splint implant consists of 2 epicardial buttons connected bya tension member that bisects the left ventricle. The Myo-splint buttons (generally 3 sets) are placed down the longLV axis to create a bilobular shape.5-7 In initial experimentalstudies Takagaki and associates6 confirmed acute LV shapechanges induced by the Myosplint and showed that thispassive device was associated with a 20% decrease in wallstress and an increase in LV ejection fraction. Cardiac

output and ventricular pressures were unaltered. These ben-efits were maintained over 30 days.5 Peak LV elastance andpreload recruitable stroke work increased from baseline to 4weeks after placing the Myosplint.7

In human subjects the Myosplint was initially tested inpatients awaiting heart transplantation. Placed just beforeremoving the heart, this passive device decreased peak LVsystolic wall stress.8 In an initial clinical series the devicewas implanted in 5 patients in New York Heart Associationclass III with dilated cardiomyopathy.8,9 Placement of theMyosplints was safely performed without early significant

Figure 4. A, Representative recordings showing LV pressure-volume loops at baseline heart failure and duringCardioClasp placement at 25% LV dimension reduction. LV contractility assessed by Emax was significantly higherby 68.0% � 21.7% with the CardioClasp (P � .01 compared with baseline heart failure). LVV, LV volume; LVP, LVpressure. B, Preload recruitable stroke work (SW) versus end-diastolic volume (EDV) relationship at baseline heartfailure (solid bullets and dotted line) and with the CardioClasp (rounded bullets and solid line). The CardioClaspsignificantly increased the preload recruitable stroke work– end-diastolic volume relationship (P � .02). MSW,Slope of the LV preload recruitable stroke work (in millimeters of mercury); EDV, LV end-diastolic volume (inmillimeters). C, Slope of the maximum dP/dt–VED (dP/dtMAX–VED) relationship (dE/dt) at baseline heart failure (solidbullets and dotted line) and with the CardioClasp (rounded bullets and solid line). The CardioClasp significantlyincreased dE/dt (in millimeters of mercury per second per milliliter), (P � .02). LVdP/dt, First derivative of the LVpressure (in millimeters or mercury per second); EDV, LV end-diastolic volume (in milliliters).



ET

adverse events. However, reconfiguration of the left ventri-cle by the Myosplints might necessitate simultaneous mitralvalve repair in patients with clinically significant incompe-tence.16

Passive Cardiac RestraintThe results of cardiomyoplasty have led to the developmentof passive devices to constrain the heart and prevent furtherventricular dilatation.17,18 One such device is the AcornCorCap CSD. The Acorn CorCap CSD is a mesh-like jacketthat is slipped around the heart and adjusted to prevent theheart from further dilatation.

Most experimental studies with the Acorn device haveinvolved induction of heart failure, placement of the device,and then, weeks to months later, examination of ventricularfunction. In an experimental study Chaudhry and cowork-ers19 produced heart failure by means of intracoronary mi-croembolization. In the control animals LV size continuedto increase, and ejection fraction decreased. In the Acorn-supported animals, LV size decreased slightly, and ejectionfraction increased. Sabbah and associates20 showed that thepercentage of myocyte shortening and the peak velocity ofshortening (dS/dt) were significantly increased in CSD-treated dogs undergoing heart failure. In a rapid ventricularpacing heart failure model, Power and colleagues21 reportedhigher LV fractional shortening and less mitral regurgitationin Acorn-support animals. In sheep 1 week after coronaryartery ligation, Pilla and associates22,23 placed the AcornCSD around the heart. At terminal study, infarct size wassignificantly smaller, and ventricular function was signifi-cantly better in the Acorn-supported animals.

In patients with idiopathic cardiomyopathy, Konertz andcoworkers24 reported a decrease in LV end-diastolic and LVend-systolic diameters and an increase in LV ejection frac-tion by using the Acorn device in patients. These benefitswere maintained at 1 year.24,25 Six patients with symptom-atic heart failure caused by ischemic cardiomyopathy were

treated surgically with the Acorn CSD and underwent cor-onary artery bypass grafting.24-26 Ventricular reconstructionwas also performed in 5 of the 6 patients. Both the LVend-diastolic and LV end-systolic diameters were signifi-cantly decreased 1 month postoperatively and did notchange significantly over the next 11 months. Mitral regur-gitation improved from a mean of 2.7 preoperatively to 1.4at 12 months, and the average LV ejection fraction alsoincreased from 27% preoperatively to 35.9% at 12 monthsafter surgical intervention.25

Limitations of the StudyDespite the fact that we used a dilated cardiomyopathyanimal model, the mechanism and dilatation in pacing-induced cardiomyopathy is different than that in the humanidiopathic cardiomyopathy and poorly understood. Furtherchronic studies are to be carried out to understand thechronic changes and effects of CardioClasp device place-ment on dilated cardiomyopathy. Currently, at Temple Uni-versity, we have started chronic studies with a pacingmodel, in which the CardioClasp is implanted for 30 days.After device implantation, the pacing parameters are con-tinued to maintain the heart failure model. We reported apreliminary report27 of chronic data as an abstract (n � 4) atthe 2002 American Society for Artificial Internal Organsmeeting. After 30 days of chronic implantation, the epicar-dial area underneath the anterior and posterior bars showedfibrous tissue formation underneath the pads. However,histologic epicardial and myocardial samples did not showany ischemic changes related to the pads. No necrosis ormyocardial injury was present at the compression sites bythe bars of the CardioClasp device. As the numbers ofanimals were few, further randomized chronic studies withsham operations are to be carried out for statistical differ-ences. These results can be used to focus on the effects ofthe CardioClasp in dilated cardiomyopathy and might becloser to those of surgical intervention for clinical dilatedcardiomyopathy.

Summary and ConclusionsThe goal of medical management of heart failure is toreduce afterload. This is usually achieved by lowering vas-cular resistance. The resulting lower arterial blood pressure

TABLE 3. Effects of the CardioClasp on LV contractilityEmax

(mm Hg/mL)MSW

(mm Hg)dE/dtmax

(mm Hg � s�1 � mL)

Baseline 1.87 � 0.47 28.4 � 11.0 10.60 � 4.58Clasp 3.22 � 1.55* 44.1 � 23.5* 18.58 � 7.42†

Emax, Slope of LV ESPVR; MSW, slope of preload recruitable stroke workversus end-diastolic volume; dE/dtmax, slope of the maximum dP/dt–VED

(dP/dtmax–VED) relationship.*P � 0.02 and †P � 0.01 when compared with baseline values.

TABLE 2. Effects of the CardioClasp on hemodynamic pa-rametersHemodynamic parameters Baseline heart failure CardioClasp

LVEDP (mm Hg) 17.5 � 3.1 17.9 � 4.3LVSP (mm Hg) 80.3 � 12.4 82.4 � 16.9Max dP/dt (mm Hg/s) 782 � 223 956 � 519Min dP/dt (mm Hg/s) �700 � 151 �786 � 291AoPsys (mm Hg) 78.4 � 12.0 79.5 � 16.6AoPdias (mm Hg) 52.0 � 4.6 58.4 � 13.4CO (L/min) 1.9 � 0.4 1.9 � 0.5

LVEDP, Left ventricular end-diastolic pressure; LVSP, left ventricular peaksystolic pressure; Max dP/dt, peak first positive derivative of the leftventricular pressure; Min dP/dt, peak first negative derivative of the leftventricular pressure; AoPsys, peak systolic aortic pressure; AoPdias, peakdiastolic aortic pressure; CO, cardiac output.



ET

can compromise blood flow to other vital organs, such as thebrain and kidney. The CardioClasp device can reduce thetension on the myocardial cells without decreasing arterialblood pressure. This leads to improved LV systolic perfor-mance. The CardioClasp significantly increased the strokework versus end-diastolic volume relationship and the slopeof the maximum dP/dt–VED(dE/dtMAX) relationship, andEmax was also significantly increased by the CardioClasp.With overall increased LV systolic indices and contractility,this passive geometric ventricular reshaping has the addedpotential for preserving the contractile mass and circumfer-ential length.

We thank Dr Milton April and Lewis Bright for their dedicatedsupport and animal care for this experiment.

References

1. Melvin DB. Ventricular radius reduction without resection: a compu-tational analysis. ASAIO J. 1999;45:160-5.

2. Kashem MA, Santamore WP, Hassan S, Crabbe DL, Margulies KB,Melvin DB. CardioClaspTM: a new passive device to re-shape cardiacenlargement. ASAIO J. 2002;48:253-9.

3. Shimizu J, Wang J, Yi GH, He K, Kashem MA, Crabbe DL, et al.Improved systolic performance by passive remodeling in experimentalheart failure. Circulation. 2000;102(suppl 18):683-3304.

4. Santamore WP, Kashem A, Hassan S, Crabbe DL, Margulies KB,Melvin DB. Left ventricular reshaping enhances contractility andresponse to increased afterload. Presented at the 8th Annual Meetingof Cardiac Technology and Transfer with STS meeting; 2002 Feb;Fort Lauderdale, Fla. Promedica International, a California Corpora-tion.

5. McCarthy P, Takagaki M, Ochiai Y, Young JB, Tabata T, Shiota T, etal. Device-based change in left ventricular shape: a new concept forthe treatment of dilated cardiomyopathy. J Thorac Cardiovasc Surg.2001;122:482-90.

6. Takagaki M, Fukamachi K, McCarthy P, Ochiai Y, Dessoffy R,Vidlund R, et al. Novel device to change left ventricular shape forheart failure treatment: device design and implantation procedure.ASAIO J. 2001;47:244-8.

7. McCarthy P, Fukamachi K, Takagaki M, Armstrong G, Young J,Schweich C, et al. Device based left ventricular shape change imme-diately reduces left ventricular volume and increases ejection fractionin a pacing induced cardiomyopathy model in dogs: a pilot study[abstract]. J Am Coll Cardiol. 2000;35:183A.

8. Fukamachi K. Myosplint: a new device to treat heart failure bychanging left ventricular shape. Proceedings of the 4th InternationalSymposium on Volume Reduction Surgery. J Card Surg. 2002:16.

9. Fukamachi K, McCarthy P, Takagaki M, Ochiai Y, Doi K, DessoffyR, et al. Myosplint improved left ventricular function by improvingsystolic elastance and maintaining diastolic compliance in a caninecardiomyopathy model. Circulation. 2001;104:II-440.

10. Shannon RP, Komamura K, Stamber BS, Manders WT, Vatner SF.Alterations in left ventricular geometry and myocardial contractility inconscious dogs with pacing induced dilated cardiomyopathy. Am JPhysiol. 1991;260:H1903-11.

11. Mirsky I, Pasipoularides A. Clinical assessment of diastolic function.Prog Cardiovasc Dis. 1990;32:291-318.

12. Spinale FG, Coker ML, Krombach SR, Mukherjee R, Hallak H,Houck WV, et al. Matrix metalloproteinase inhibition during thedevelopment of congestive heart failure: effects on left ventriculardimensions and function. Circ Res. 1999;85:364-76.

13. Sagawa K. The end-systolic pressure-volume relation of the ventricle:definition, modifications, and clinical use. Circulation.1981;63:1223-7.

14. Glower DD, Spratt JA, Snow ND, Kabas JS, Davis JW, Olsen CO, etal. Linearity of the Frank-Starling relationship in the intact heart: the

concept of preload recruitable stroke work. Circulation. 1985;71:994-1009.

15. Little WC. The left ventricular dP/dt max end-diastolic volume rela-tion in closed-chest dogs. Circ Res. 1985;56:808-15.

16. Schenk S, Reichenspurner H, Groezner JG, Boehm DH, Schirmer J,Scheidt WV, et al. Myosplint implantation and ventricular shapechanges in patients with dilative cardiomyopathy—first clinical expe-rience. J Heart Lung Transplant. 2001;20:217.

17. Patel HJ, Polidori DJ, Pilla JJ. Stabilization of chronic remodeling byasynchronous cardiomyoplasty in dilated cardiomyopathy: effects of aconditioned muscle wrap. Circulation. 1997;96:3665-71.

18. Kass DA, Baughman KL, Pak PH, Cho PW, Levine HR, Gardner TJ,et al. Reverse remodeling from cardiomyoplasty in human heart fail-ure. External constraint versus active assist. Circulation. 1995;91:2314-8.

19. Chaudhry PA, Mishima T, Sharov VG, Hawkins J, Alferness CA,Paone G, et al. Passive epicardial containment prevents ventricularremodeling in heart failure. Ann Thorac Surg. 2000;70:1275-80.

20. Sabbah HN, Chaudhry PA, Paone G, Mishima T, Alferness C. Passiveventricular constraint with the Acorn prosthetic jacket prevents pro-gressive left ventricular dilation and improves ejection fraction in dogswith moderate heart failure [abstract]. J Am Coll Cardiol. 1999;33:207A.

21. Power JM, Raman J, Dornom A, Farish SJ, Burrell LM, Tonkin AM,et al. Passive ventricular constraint amends the course of heart failure.Cardiovasc Res. 1999;44:549-55.

22. Pilla JJ, Blom AS, Brockman DJ, Bowen F, Yuan Q, Giammarco J, etal. Ventricular constraint using the Acorn cardiac support device(CSD) limits infarct expansion. Circulation. 2001;104:II-480.

23. Pilla JJ, Brockman DJ, Blom AS, Bowen F, Yuan Q, Gorman JH, etal. Prevention of dilation using the Acorn cardiac support device(CSD) results in reverse remodeling and improvement of function.Circulation. 2001;104:II-440.

24. Konertz WF, Shapland JE, Hotz H, Dushe S, Braun JP, Stantke K, etal. Passive containment and reverse remodeling by a novel textilecardiac support device. Circulation. 2001;104(suppl I):I-270-5.

25. Konertz WF, Kleber FX, Dushe S, Hotz H, Stantke K. Efficacy trendswith the Acorn Cardiac Support Device: One year follow-up. Circu-lation. 2001;104:II-357.

26. Raman J, Power JM, Buxton BF, ALferness CA, Hare D. Ventricularcontainment as an adjunctive procedure in ischemic cardiomyopathy:early results. Ann Thorac Surg. 2000;70:1124-6.

27. Kashem A, Hassan S, Santamore WP, Crabbe DL, Margulies KB,Goldman BI, et al. Chronic experimental heart failure: CardioClaspTM

can maintain improved myocardial function [abstract]. J ASAIO. 2002;48:161.

DiscussionDr W. Randolph Chitwood (Greenville, NC). Dr Kashem, I

think you have done an eloquent job in studying physiologicallythe changes associated with these new devices, which we hope willprovide great improvement in hemodynamics for patients withthese terribly dilated ventricles. They really had no hope in thepast. We congratulate you on your work. I have several question-s.The first question is, how is this device incorporated into hearttissue? Did you do any studies looking at the incorporation of thedevice into the myocardium?

Dr Kashem. We checked with the histology from chronic andacute studies, and we did not find any problem with the histology,and there was no significant abrasion incorporating the device overto the myocardium. Histologically, it did not show any grossabnormalities of the myocardium.

Dr Chitwood. After this device is working and active, does itchange papillary muscle function, tip distance, or angulation withthe leaflets through the chords? Was there any mitral insufficiencyseen in these patients?



ET

Dr Kashem. No, we did not present those date here, but at theInternational Society for Heart and Lung Transplantation meeting,I presented the mitral regurgitation. We performed transesophagealechocardiography in all the animal studies, which we did not showhere, and I found there are no changes in mitral regurgitation.Mostly in heart failure we found that mitral regurgitation wasgrade 1 to 2 in most cases, and we found there was no significantalteration of the mitral regurgitation in those cases.Regardingpapillary muscle, I did not see any implications with the Cardio-Clasp device.

Dr Chitwood. Do you think this device could actually be usedin a different way to reorient papillary muscles in patients whohave ischemic mitral regurgitation or cardiomyopathic mitral re-gurgitation? Can you reorient the chordae? Could this devicedecrease mitral insufficiency?

Dr Kashem. We did not investigate the means of the place-ment. We did investigate several ways to place the device, and wedid not find any significant changes. But maybe in the future wecan see whether there are any significant effects on the papillarymuscle or mitral regurgitation by placing the device.

Dr Chitwood. How does this device differ from the Myocorand Acorn devices in its efficacy?

Dr Kashem. The basic theory for the Myocor and CardioClaspis the same—to reduce wall stress—but the important thing here isthat it is a non-blood contact device for CardioClasp. It does notpenetrate the left ventricle. The other one has a bilobular shape,and the CardioClasp does the same thing. Compared with theAcorn device, the CardioClasp device has an immediate effect oncontractility and an immediate effect on reduction of the wallstress and tension on the myocardial cells. It should have a reverseremodeling in long-term cases; we did not study those cases. Iguess a 3-month or 6-month study could answer the question ofwhether there is a reverse remodeling at the myocyte level.

Dr Chitwood. It would be interesting to compare these devicesas to their efficacy and efficiency using your hemodynamic mod-els.

Dr Kashem. Yes.Dr Chitwood. What is the ideal target patient population for

these devices?Dr Kashem. Mostly patients undergoing dilated cardiomyop-

athy of ischemic or nonischemic origin. The CardioClasp devicecould be placed in New York Heart Association class III or IVpatients. But we still have to go further on the inclusion-exclusioncriteria. We did not decide that.



ET

Left ventricular reshaping: Effects on the pressure-volume relationship

Documents

Transcript of Left ventricular reshaping: Effects on the pressure-volume relationship